Obesity resists simple treatments, especially in today's environment which promotes unhealthful eating, sedentary behaviors, and poor sleep habits. Early childhood is a developmentally plastic period during which the right environmental and developmental influences may entrain healthful weight trajectories for life. Preventing obesity in early childhood cannot be achieved by working in single disciplines, but rather warrants a broad systems perspective to frame the problem, identify critical leverage points of influence, and take effective action. Whole communities represent loci in which multiple sectors, levels, and key leverage points converge. The overall goal of this US-Australian collaborative project is to advance intervention science by addressing what works, for whom and under what circumstances? To achieve this goal, we will use a novel, iterative, approach that combines computational systems science modeling with empirical information from two completed and two new whole-of-community intervention programs. We will develop, test, and refine our computational agent-based model through repeated and bi-directional interaction with data from these interventions. Through these efforts we aim not only to design and refine our own obesity prevention efforts in young children, but also to create a flexible agent-based model with wide applicability to aid future efforts. Beginning with dynamic hypotheses from our preliminary systems mapping work and the literature, we will apply and test an initial model using already collected data from 2 completed interventions that successfully changed behavior and BMI: 1.) Shape Up Somerville (Mass.), a community-wide, multi-environmental and policy intervention among school-age children; and 2.) Romp & Chomp, a whole-of-community obesity prevention demonstration project for preschoolers in one city in the state of Victoria, Australia. Next, we will iteratively refine the model using fine grained data drawn from participatory group model building exercises in both Somerville and Australia, and process data from a new state-wide quasi-experimental trial of childhood obesity intervention trial in Victoria. We will then apply the improved model and insights generated from this first set of studies to design, plan and conduct a new pilot intervention, Shape Up Under-5, a re- designed version of Shape Up Somerville targeting underserved children age 0-5 years and their primary caregivers. At the end of the proposed project, we will test and calibrate the final model with outcome data from the multi-year trial in Victoria and the Somerville pilot. To achieve these aims we have assembled a trans-disciplinary research team with extensive expertise in obesity prevention, computational systems modeling, community-based research, and implementation science. The impact of this project lies in the potential to inform new and sustainable strategies to reduce the burden of childhood obesity and its consequences in the US and abroad.